Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Getting 1D and 2D results to match
Posted 29 oct. 2018, 16:00 UTC−4 Heat Transfer & Phase Change, Geometry, Parameters, Variables, & Functions 6 Replies
Please login with a confirmed email address before reporting spam
I am trying to solve a 2D planer and 1D line heat transfer model. The 2D model has a heat generating source in the middle and is surrounded by various materials with different thermal conductivities. The outer face is a fixed temperature (subject to coolant, this will be adjusted later). The 2D model gives us results that are similar ( ~2-3%) to the results we get with Abaqus. We are on a demo license for COMSOL to see if we want to purchase it and use it to replace the analysis that we currently do with Abaqus. The 2D model is fine so far. However, we also use 1D models in Abaqus to quickly evaluate parameter changes and so want to be sure that COMSOL povides similar results. COMSOL does not. The resulting peak temperature in our source is approximately twice the tempereature as we get in the 2D (1000+ K compared to ~400 K). Is there any reason that the 1D would solve the exact same parameters so differently?
Please login with a confirmed email address before reporting spam
Dear Geoffrey, Thanks for sharing your feedback. It is difficult to be sure without checking the mph file but I suspect that either the Cross sectional area or the Cross sectional perimeter values are not set to the desired values. You can check and edit them in the physics interface node ("Heat Transfer in Fluids" for example). Feel free to share you mph file or to contact the support if the results still don't match.
Best regards,
Nicolas
nicolas@comsol.com
Please login with a confirmed email address before reporting spam
Nicolas,
Thank you for the response. I'll try and play around with the area and perimeter. I've attached the 1D mph file. The 2D was too big to attach or I would have included that for comparison. Cheers,
-Boone
Attachments:
Please login with a confirmed email address before reporting spam
Nicolas,
I've tried playing around with the Ac and Pc, but it makes no difference to the results. I've tried looking up tutorials on 1D models for heat transfer but all there seems to be is the radiant heating model and nothing in conductive heat transfer in a solid.
-Boone
Please login with a confirmed email address before reporting spam
Hi Geoffrey/Boone,
You should be able to make your 2D file much smaller by clearing all solutions (Study ribbon). If it's still too large to post, you can also clear all meshes (Mesh ribbon).
Best,
Jeff
-------------------Jeff Hiller
Please login with a confirmed email address before reporting spam
Hi again Geoffrey,
From the parameters names I'm wondering if what you are looking for is named "1D axisymetric geometry" in COMSOL Multiphysics. You can see a 1D geometry as a rod with a given cross section. In a 1D geometry a segment represent the rod in its length direction. A 1D axisymetric geometry is different, it represent a disk. In attached MPH-file you'll see the comparison between a 1D geoemtry and a 1D axisymetric geometry. I noticed that for second case the maximum temperature is about 550K.
I hope it helps,
Nicolas
Attachments:
Please login with a confirmed email address before reporting spam
I think I figured out my problem. My 2D is a planar model (essentially a cross section of the middle of my model where it should be the hottest). I rebuilt the 1D model as a 1D axisymmetric model and I got pretty much the same answer.
-Boone